Managing bulk scanning and ingestion

If you have got millions of paper docs to be digitized and ingested into UCM with the required metadata, Oracle Document Capture can be effectively utilized for this cause.

This post provides some insight into how to accomplish this and how to manage millions of docs ingested in lesser time. Various instances of Import Server, Recognition Server and Commit Server can be setup and executed simultaneously to speed up the post-scanning processing. Once the paper docs are scanned and dumped at a location, further processing of importing the scanned images to create ODC batches, recognizing the metadata values through a barcode, applying the read values to the indexing fields and finally committing the indexed batch to UCM can be fully automated.

You need to understand the following limitations and caveats of such a setup:

1) Licenses. You would need multiple licenses of the product. Do a thorough analysis and arrive at a number of licences of the ODC that you would need.

2) Import Server. Individual Import Server would need a separate folder where it can listen for new scanned images to be imported. Unlike other modules of the ODC stack, multiple Import servers cannot be configured to listen to one folder.

3) Recognition Server. Just setup to listen to the same database as other modules. No other configuration required to make it run in parallel with other module like Commit Server. The Recognition process will lock the current processing batch, so that no other component (other Recognition or Commit) can make any update on this batch. Dont setup the commit here, configure a separate set of Commit Servers.

4) Commit Server.Setup similar to Recognition, Can run in parallel with other modules. Keep it installed separately than other modules of Import and Recognition, as it would take most of the time of your batch processing. Recognition happens pretty quick. Commit takes most of the time.

5) Database. As part of the ODC setup, this will be configured in all the modules.All these components can be pointing to the same DB simultaneously.

6) File Store. If all the modules are installed on different desktops, only requirement for them to function without any issues is a shared file location.Scanners will dump scanned images to this share, Import will pickup scanned images and import them to create batches in centralized db.

7) Admin Server. Have a separate ODC desktop configured for the ODC Admin to login and configure the system, all other desktops with the different modules can be executed with a non-admin user. So that the db , batch or any other config does not get edited/updated/messed-up knowingly or un-knowingly.

This would help you in maximizing the processing power of ODC components.